plldesign
Newbie level 4
I am doing hand calculation on the phase noise of ring oscillator. The technology is 0.18 um.
I am using Hajimiri's theory, and follow his thesis "Jitter and Phase Noise in Electrical Oscillators". I used the formula of (5.17).
(5.17) reads as:
L(dw) = (8/(3*eta))*(kT/P)*(Vdd/Vchar)*(f0^2/df^2)
The frequency offset I use dw = 1MHz
f0 = Ueff*Weff*Cox*DV^2/(8*eta*N*L*qmax)
qmax = 1.8*Cox*(Wp+Wn)*L*2/3
I use eta as 0.85, and gamma = 2.
I want to know whether my formula of qmax is correct, i use 2/3 in that formula by following the book of Thomas Lee (forget the details). i choose eta to fit the simulation result.
the hand calculation phase noise is -90dBc/Hz and the simulated one (by Cadence) is -88dBc/Hz. Seems fit well, however, i choose eta to fit that value and f0 (which also fit well)
i just want to hear your opinion about whether my calculation is correct, and whether the value of eta and gamma is correct.
thanks.
I am using Hajimiri's theory, and follow his thesis "Jitter and Phase Noise in Electrical Oscillators". I used the formula of (5.17).
(5.17) reads as:
L(dw) = (8/(3*eta))*(kT/P)*(Vdd/Vchar)*(f0^2/df^2)
The frequency offset I use dw = 1MHz
f0 = Ueff*Weff*Cox*DV^2/(8*eta*N*L*qmax)
qmax = 1.8*Cox*(Wp+Wn)*L*2/3
I use eta as 0.85, and gamma = 2.
I want to know whether my formula of qmax is correct, i use 2/3 in that formula by following the book of Thomas Lee (forget the details). i choose eta to fit the simulation result.
the hand calculation phase noise is -90dBc/Hz and the simulated one (by Cadence) is -88dBc/Hz. Seems fit well, however, i choose eta to fit that value and f0 (which also fit well)
i just want to hear your opinion about whether my calculation is correct, and whether the value of eta and gamma is correct.
thanks.
